【摘要】 研究提出利用声信号有效值进行内积运算识别白噪声声源的方法,用传声器阵列测得声压信号有效值组成声压有效值向量,与虚拟点声源声压有效值向量做内积,搜索内积模的最大值,根据柯西施瓦兹不等式确定声源位置。建立相应的识别步骤,对白噪声声源进行仿真识别,分析测量误差与测量距离对识别效果的影响,通过增大传声器布置间距改善测量距离较远时的识别效果。在能够识别单个声源的基础上研究多个声源的识别,主要针对两个声源间距近的情况,提出结合施密特正交对其进行识别,识别结果准确。进行相应的实验验证,证明方法的有效性。由于该方法不需要使用相位信息,具有测量方便,对实验设备要求低等特点。更多还原

【Abstract】 This research proposes a method to identify the source of white noise by using the effective value of the acoustic signal to perform inner product operations. The effective value of the sound pressure signal measured by the microphone array is formed into a sound pressure effective value vector. Search for the maximum value of the inner product mode, and determine the sound source position according to the Cauchy Schwartz inequality. Establish the corresponding identification steps and simulate the identification of the white noise source. Analyzed the influence of measurement error and measurement distance on the recognition effect, and improved the recognition effect when the measurement distance was farther by increasing the distance between microphones. Based on the recognition of a single sound source, the recognition of multiple sound sources is studied. For the situation where the distance between two sound sources is close, a combination of Schmidt orthogonal recognition is proposed, and the recognition result is accurate.Experimental verification proves the effectiveness of the method. Since this method does not need to use phase information, it has the characteristics of convenient measurement and low requirements for experimental equipment.更多还原